The long term objective of this work is to predict the structure and energetics of protein-ligand non-covalent and covalent interactions. By comparison with appropriate solution phase interactions and binding and catalysis using organic biomimetic models, the nature of enzyme catalysis will be better understood. This understanding should lead to new ideas in site-directed mutagenesis approaches to enzyme design, as well as new biomimetic models to match the catalytic efficiency of protein enzymes. These studies will also lead to new ideas on the design of non-covalent and covalent enzyme inhibitors as transition state analogs. Since such inhibitors are of great potential medical importance (e.g., angiotensin converting enzyme inhibitors to combat hypertension), the studies proposed here could have an important impact on human medicine. We propose to develop a combined quantum/molecular mechanical methodology for the study of protein-ligand interactions, in which ab initio methods are used for the chemically reacting groups, combined with a molecular mechanical representation of both protein and water solvent. The solvent will be represented explicitly with accurate analytical potentials. The focus of the studies here will be on the serine and sulfhydryl proteases, on triose phosphate isomerase, as an example of a phosphate-binding enzyme, on staphylococcal nuclease, as an example of phosphate hydrolase and on adenylate kinase, as an example of a phosphate kinase. This particular set of proteins are being studied not only because their properties provide an important and stringent test for the theoretical methodologies, but also because of their physiological importance.